Protocol Online logo
Top : New Forum Archives (2009-): : SDS-PAGE and Western Blotting

Antibody Characterization - KO vs. WT or Heterologous overexpression (Jul/07/2009 )

Hello,

I am having a marathon ordeal of showing that a custom made antibody I'm using is legitimate.

So far we have positive results on peptide preabsorption on both IHC and WB, heavy detection of the appropriate band after heterologous overexpression in HEK cells, a lack of staining with mock transfection controls, lack of staining in IHC with secondary antibody alone, and a positive band at the right weight after IP and no band at the right weight with the mock transfection IP. There are lower bands present that may represent cleavage products or alternative splicing products. It is also generally believed that nearly all commercial antibodies available are suspect (hence the custom made version)

Here is where it gets gray: In a KO animal (gift), we do not see the molecular weight band (good!), but see a prominent band at approx 27kD. Has anyone had this experience when looking at an embryonic KO animal to test their antibody and seeing non-specific staining of lower molecular weight bands appear?

Does this mean the antibody is no good?

Any insight would be appreciated, we're looking into mass spec and asking for perfused KO brain tissue to look at what we see on that.

Thanks!

-Neurodarcy-

Neurodarcy on Jul 8 2009, 05:46 AM said:

Hello,

I am having a marathon ordeal of showing that a custom made antibody I'm using is legitimate.

So far we have positive results on peptide preabsorption on both IHC and WB, heavy detection of the appropriate band after heterologous overexpression in HEK cells, a lack of staining with mock transfection controls, lack of staining in IHC with secondary antibody alone, and a positive band at the right weight after IP and no band at the right weight with the mock transfection IP. There are lower bands present that may represent cleavage products or alternative splicing products. It is also generally believed that nearly all commercial antibodies available are suspect (hence the custom made version)

Here is where it gets gray: In a KO animal (gift), we do not see the molecular weight band (good!), but see a prominent band at approx 27kD. Has anyone had this experience when looking at an embryonic KO animal to test their antibody and seeing non-specific staining of lower molecular weight bands appear?

Does this mean the antibody is no good?

Any insight would be appreciated, we're looking into mass spec and asking for perfused KO brain tissue to look at what we see on that.

Thanks!

I agree, all looks good when you test your antiserum against recombinant protein. I assume the polyclonal was made against the recombinant protein:

In the IHC did you express in mouse or human cells (if human, the 27kD reactivity may only be seen with mouse, hence the neg mock transfection.

More importantly, do you see the 27KD band in westerns with normal (non-KO) embryos?

-klinmed-


I agree, all looks good when you test your antiserum against recombinant protein. I assume the polyclonal was made against the recombinant protein:

In the IHC did you express in mouse or human cells (if human, the 27kD reactivity may only be seen with mouse, hence the neg mock transfection.

More importantly, do you see the 27KD band in westerns with normal (non-KO) embryos?


The antibody was raised against a peptide matching the human protein sequence in a rabbit host. The sequence is identical to mouse in terms of the peptide sequence. overall there is pretty much >99% homology.

For IHC we are using perfused mouse brain tissue and for recombinant expression we are using HEK cells. The recombinant protein used for the vector was also made using the human gene sequence.

In normal mice, we do see the 27kD band as well. At first we used to see a band at the bottom where the dye front was running off the gel and where 25kD normally ended. Now it seems that separating out these smaller bands even farther shows this 27kD band and not the dye front blob.

The correct molecular weight for the protein is 140.

-Neurodarcy-

Neurodarcy on Jul 8 2009, 05:46 AM said:

Hello,

I am having a marathon ordeal of showing that a custom made antibody I'm using is legitimate.

So far we have positive results on peptide preabsorption on both IHC and WB, heavy detection of the appropriate band after heterologous overexpression in HEK cells, a lack of staining with mock transfection controls, lack of staining in IHC with secondary antibody alone, and a positive band at the right weight after IP and no band at the right weight with the mock transfection IP. There are lower bands present that may represent cleavage products or alternative splicing products. It is also generally believed that nearly all commercial antibodies available are suspect (hence the custom made version)

Here is where it gets gray: In a KO animal (gift), we do not see the molecular weight band (good!), but see a prominent band at approx 27kD. Has anyone had this experience when looking at an embryonic KO animal to test their antibody and seeing non-specific staining of lower molecular weight bands appear?

Does this mean the antibody is no good?

Any insight would be appreciated, we're looking into mass spec and asking for perfused KO brain tissue to look at what we see on that.

Thanks!

So, with normal mice you see a 140 kD and 27 kD band and with the KO only the 27 kD band.

Sounds like your anti-peptide antiserum cross reacts with a 27 kD protein present in both WT and KO mice.
You need to find out if the anti-27 kD reactivity in you antibody reacts with the free peptide (do not use the conjugate that you made for immunization).
I would do this by absorbing a few ul of you antiserum with a large excess of free peptide. Do this overnight at 4 oC. Then dilute and do western. Does both the 140 AND 27 bands become weaker?

If they do, you have crossreactivity with the anti-peptide antibodies present in the antiserum. However, if the 27 kD band remains, it would suggest you may also have some antibodies to say the carrier protein used for immunization. in this case you will be able to immunoaffinity purify you antibodies on immobilized peptide columns.

Hope this helps

-klinmed-

klinmed on Jul 9 2009, 04:18 AM said:

Neurodarcy on Jul 8 2009, 05:46 AM said:

Hello,

I am having a marathon ordeal of showing that a custom made antibody I'm using is legitimate.

So far we have positive results on peptide preabsorption on both IHC and WB, heavy detection of the appropriate band after heterologous overexpression in HEK cells, a lack of staining with mock transfection controls, lack of staining in IHC with secondary antibody alone, and a positive band at the right weight after IP and no band at the right weight with the mock transfection IP. There are lower bands present that may represent cleavage products or alternative splicing products. It is also generally believed that nearly all commercial antibodies available are suspect (hence the custom made version)

Here is where it gets gray: In a KO animal (gift), we do not see the molecular weight band (good!), but see a prominent band at approx 27kD. Has anyone had this experience when looking at an embryonic KO animal to test their antibody and seeing non-specific staining of lower molecular weight bands appear?

Does this mean the antibody is no good?

Any insight would be appreciated, we're looking into mass spec and asking for perfused KO brain tissue to look at what we see on that.

Thanks!

So, with normal mice you see a 140 kD and 27 kD band and with the KO only the 27 kD band.

Sounds like your anti-peptide antiserum cross reacts with a 27 kD protein present in both WT and KO mice.
You need to find out if the anti-27 kD reactivity in you antibody reacts with the free peptide (do not use the conjugate that you made for immunization).
I would do this by absorbing a few ul of you antiserum with a large excess of free peptide. Do this overnight at 4 oC. Then dilute and do western. Does both the 140 AND 27 bands become weaker?

If they do, you have crossreactivity with the anti-peptide antibodies present in the antiserum. However, if the 27 kD band remains, it would suggest you may also have some antibodies to say the carrier protein used for immunization. in this case you will be able to immunoaffinity purify you antibodies on immobilized peptide columns.

Hope this helps


Very interesting,

So the statement you made about the possibility of crossreactivity with the anti-peptide antibodies present in the antiserum, are you saying that the antibody is recognizing other antibodies from the serum? Preabsorption with peptide does get rid of both bands, but I may be missing your difference between free peptide and the conjugated version. If I am understanding correctly, the conjugate peptide made for immunization basically has an extra cysteine bridge and an acetylated tag on it while the free peptide is basically just the string of amino acids. If thats the case, then the preaborption we have done is with the free peptide and does block both bands.

If its just antibody cross reactivity, that would be great. However, this antibody is actually already affinity purified. Does that explain anything new?

Thank you for all your help.

-Neurodarcy-

To me it seems that your KO mouse is not a complete knockout of the gene (which would be unusual anyway), but enough of the gene is lost such that the protein is non-functional. I suspect you are getting expression of the mutant gene as a smaller product caused by the disruption of the coding sequence, and this is the 27 kDa band you are seeing.

-bob1-

Neurodarcy on Jul 9 2009, 04:45 PM said:

klinmed on Jul 9 2009, 04:18 AM said:

Neurodarcy on Jul 8 2009, 05:46 AM said:

Hello,

I am having a marathon ordeal of showing that a custom made antibody I'm using is legitimate.

So far we have positive results on peptide preabsorption on both IHC and WB, heavy detection of the appropriate band after heterologous overexpression in HEK cells, a lack of staining with mock transfection controls, lack of staining in IHC with secondary antibody alone, and a positive band at the right weight after IP and no band at the right weight with the mock transfection IP. There are lower bands present that may represent cleavage products or alternative splicing products. It is also generally believed that nearly all commercial antibodies available are suspect (hence the custom made version)

Here is where it gets gray: In a KO animal (gift), we do not see the molecular weight band (good!), but see a prominent band at approx 27kD. Has anyone had this experience when looking at an embryonic KO animal to test their antibody and seeing non-specific staining of lower molecular weight bands appear?

Does this mean the antibody is no good?

Any insight would be appreciated, we're looking into mass spec and asking for perfused KO brain tissue to look at what we see on that.

Thanks!

So, with normal mice you see a 140 kD and 27 kD band and with the KO only the 27 kD band.

Sounds like your anti-peptide antiserum cross reacts with a 27 kD protein present in both WT and KO mice.
You need to find out if the anti-27 kD reactivity in you antibody reacts with the free peptide (do not use the conjugate that you made for immunization).
I would do this by absorbing a few ul of you antiserum with a large excess of free peptide. Do this overnight at 4 oC. Then dilute and do western. Does both the 140 AND 27 bands become weaker?

If they do, you have crossreactivity with the anti-peptide antibodies present in the antiserum. However, if the 27 kD band remains, it would suggest you may also have some antibodies to say the carrier protein used for immunization. in this case you will be able to immunoaffinity purify you antibodies on immobilized peptide columns.

Hope this helps


Very interesting,

So the statement you made about the possibility of crossreactivity with the anti-peptide antibodies present in the antiserum, are you saying that the antibody is recognizing other antibodies from the serum? Preabsorption with peptide does get rid of both bands, but I may be missing your difference between free peptide and the conjugated version. If I am understanding correctly, the conjugate peptide made for immunization basically has an extra cysteine bridge and an acetylated tag on it while the free peptide is basically just the string of amino acids. If thats the case, then the preaborption we have done is with the free peptide and does block both bands.

If its just antibody cross reactivity, that would be great. However, this antibody is actually already affinity purified. Does that explain anything new?

Thank you for all your help.

Hi again!

I may have not been too clear in my last post:

When you immunized you must have used your peptide coupled to a carrier protein (eg BSA, KLH?).

Thus the immunized animal would have made antibodies to BOTH your peptide AND the carrier protein. In theory you could get extra bands on a western due to antibodies in your polyclonal that are made in response to epitopes present on your carrier protein.

For example, if someone used peptide-BSA conjugates to immunize they would get a mix of anti-peptide and anti-BSA antibodies. If they did a western with cell grown in FCS they would see extra bands due to contaminating BSA (or BSA fragments or albumins from other species) being present in their cell extracts. This is one reason that KLH is often used for carrier (The keyhole limpet is phylogenetically far removed from mammals).

This is why I suggested you absorb the antiserum with unconjugated peptide. If you had reactivity due to your carrier protein you would only see the specific band disappear.

You mentioned that you Ab is affinity purified (I assume on a column of immobilized unconjugated peptide). This would have removed anti-carrier protein antibodies.

Therefore your 27 kD band is almost certainly a cross-reaction.


Bob 1 suggested that the 27 KD band could result from a partial knockout. This seems very unlikely since you see it in your wild-type (normal) mice.

Hope this helps

-klinmed-

Make sure what kind of knockout-mouse you're using. Is it an "classical" knockout (homologous recombination of a neo-vector into genome ) ? Is it maybe a "gene trap"-mouse line ? In such case, a vector with an artificial splice integrates into one intron of the gene of interest. A (in most cases loss-of-function) fusion-protein, lacking all AS after the artificial splice site but including the reporter-construct is synthesized. Such a protein should be degraded in most cases rapidly in lysosomes or the proteasome, but could be sufficient to be detected in WB or IHC.

-markusda-

markusda on Aug 1 2009, 04:42 AM said:

Make sure what kind of knockout-mouse you're using. Is it an "classical" knockout (homologous recombination of a neo-vector into genome ) ? Is it maybe a "gene trap"-mouse line ? In such case, a vector with an artificial splice integrates into one intron of the gene of interest. A (in most cases loss-of-function) fusion-protein, lacking all AS after the artificial splice site but including the reporter-construct is synthesized. Such a protein should be degraded in most cases rapidly in lysosomes or the proteasome, but could be sufficient to be detected in WB or IHC.


This mouse line definitely originated from the "classical" knockout design. A nuclear localized lacZ gene sequence was also inserted in front of the neo-vector cassette before recombination into the genome.

The gene trapping technique sounds very interesting, I haven't had any experience with this kind of mouse.

Thanks!

-Neurodarcy-